ntt123 commited on
Commit
cd75eda
1 Parent(s): b46cabd

Create app.py

Browse files
Files changed (1) hide show
  1. app.py +42 -0
app.py ADDED
@@ -0,0 +1,42 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ from pathlib import Path
2
+ from vietTTS.hifigan.mel2wave import mel2wave
3
+ from vietTTS.nat.text2mel import text2mel
4
+ from vietTTS import nat_normalize_text
5
+ import numpy as np
6
+
7
+ import os
8
+ os.system("gdown --id 16UhN8QBxG1YYwUh8smdEeVnKo9qZhvZj -O duration_latest_ckpt.pickle")
9
+ os.system("gdown --id 1-8Ig65S3irNHSzcskT37SLgeyuUhjKdj -O acoustic_latest_ckpt.pickle")
10
+ os.system("gdown --id 19cRNDC6IrHFAAE4U9I7K0mzLMgPsi5zb -O hk_hifi.pickle")
11
+
12
+
13
+ def text_to_speech(text):
14
+ text = nat_normalize_text(text)
15
+ mel = text2mel(
16
+ text,
17
+ "lexicon.txt",
18
+ 0.2,
19
+ Path("acoustic_latest_ckpt.pickle"),
20
+ Path("duration_latest_ckpt.pickle")
21
+ )
22
+ wave = mel2wave(mel, Path("config.json"), Path("hk_hifi.pickle"))
23
+ return (wave * (2**15)).astype(np.int16)
24
+ import gradio as gr
25
+
26
+
27
+ def speak(text):
28
+ y = text_to_speech(text)
29
+ return 16_000, y
30
+
31
+ title = "vietTTS"
32
+ description = "A vietnamese text-to-speech demo."
33
+
34
+ iface = gr.Interface(
35
+ fn=speak,
36
+ inputs="text",
37
+ outputs="audio",
38
+ title = title,
39
+ description=description
40
+ )
41
+
42
+ iface.launch()